How's this for an LN2 project?

[mavalpha]

I'm thinking of trying out some extreme OC'ing later this summer. I haven't really sat down and sketched the details out yet, but here's the general idea of what I'm thinking:

Let me start by saying this is NOT a modding project, so I don't care how neat or chaotic the case looks. The whole case will be made of plexiglass/Lexan or something similar, with caulking around the seams to make it airtight. I'll have a Rival VS110 Vacuum Sealer* hooked up to suck out the air, after running some cables through the walls for power, video, keyboards, etc. The drive(s) will be SATA to make routing and sealing easier- I just want to OC as far as I can, so the very fact that it's LN2 tells you how much I care about running anything major on it. I don't even need an optical drive, so long as a hard drive is preinstalled with all the apps I'll need.

Now for the fun part: the three pipes in the pic are filled with LN2, to allow for easy access and some awesome fogging effects. (I am not volt-modding my video card, so I probably won't use the one on the video card.) I decided to vacuum-seal the case rather than worry about condensation, but then remembered about RAM, MOSFETs, etc. After thinking about it, I ruled out a separate WC loop and instead decided to leave some room for Dry Ice blocks inside the case. Ideally the vacuum would keep it solid indefinitely, but I'm counting on the imperfections of my design, and the warm surface of the case, to sublimate the DI into ultra-cold vapors inside the void. This should be enough for at least an hour of benching, since the vapors will keep the video, RAM, power circuitry, and other ICs cool enough for benchmarking. I should probably mention that a lot of this stuff will need LOTS of cooling:
I'll be running BH-5 with around 4V, maybe a bit more by the time I mod my power supply.
I have three procs to choose from:
One is my Newcastle 3400+, which I've stabilized at 2.7 for now. The mobo is my infamous DFI, which I just replaced thanks to the awesome service at NewEgg. This one will require a DDR Booster, which I already have. If I mod my 5V line to 6V, I can use the Booster to run up to 4.8V or 5V through my RAM, though I don't intend to go anywhere near that high.
The second would be a 939 chip, probably a Venice 3200+. I don't know anything about it yet, since this is the chip I'll get if one my friends buys my current rig. Again, the DFI board- without any doubts or reservations. No Booster here, but again- 5V line modded to 6V if I can.
Last but not least, I'm buying a P-M 1.5 Banias core from someone here on the forums (thanks, Napalm00!) I've got an Asus P4P800 Deluxe, and a CT-479 will run me about $50. I probably will NOT use this rig, because OCZ says the Booster will only take me to 3.4V on this board. Even presuming I mod my PSU, that still won't take me to 4V, and I need to get that high for my RAM to shine. I really want to use my BH-5 for this, since it's the one time I won't have to worry about cooling it.

If you guys see any potential prolems or improvements, please let me know! I'm new to extreme cooling, and I want to make sure this goes without a hitch! I'm putting hundreds of dollars' worth of stuff on the line, so I'd really rather not watch it go up in smoke.

Also, I realized afterwards that the CO2 vapors will probably be enough for my chipset- even presuming a lowly 10x multiplier on my A64, I only need to run a 300HTT. 275 is no problem now, and I'm using the stock DFI heatsink without so much as an extra fan. I probably will dump the LN2 tower for the NB as well.

TIA,

Shaggy



*Sorry, I couldn't find a product page for this product. If anyone else can find it, please post!

Edit:
Oh, sorry- this should have gone under EXTREME Cooling Solutions. Feel free to relocate if you want, and sorry again.

 

[Tim_axe]

You'll need the HDD outside of the vacuum enclosure because they can't operate under a vacuum. That gets rid of one thing to cool inside of it...

Also, is that vacuum sealer up to the task of pumping air out of a solid like that? And can the enclosure hold up under the vacuum?


Maybe instead of trying to take the air out you can replace the air with something else that doesn't condense to form water. Since you'll have LN2 involved in this project, how about evaporating some of that into a gas to replace the air with? Nitrogen is heavier than air, so you'd be able to pour it into the case if you engineer it correctly. Then seal/cap it, and it would be impossible for water to condense.

Then again I haven't worked with LN2 myself, so I just took some time to try and learn a few things about it...


Also be sure to operate with windows open. If you get a lot of LN2 you could displace all of the oxygen in your room and suffocate. Cheers

 

[mavalpha]

You'll need the HDD outside of the vacuum enclosure because they can't operate under a vacuum. That gets rid of one thing to cool inside of it...

Thought of that, it's why I went with SATA cables- they'll be easier to caulk into the side panel than PATA. Rounder, narrower, and generally more maneuverable.

Also, is that vacuum sealer up to the task of pumping air out of a solid like that?

Yep, my preliminary tests were satisfactory. I chose the model because my mom bought one for her kitchen, and I spent the first two days getting an idea of how strong it was.

And can the enclosure hold up under the vacuum?

Thought of that too, spent the past half a day thinking whether to use a steel/aluminum skeleton, or design a mounting for the heatsink(s) so that they would actually support the top, bottom and sides (without transferring that load to the mobo, obviously).

Maybe instead of trying to take the air out you can replace the air with something else that doesn't condense to form water. Since you'll have LN2 involved in this project, how about evaporating some of that into a gas to replace the air with? Nitrogen is heavier than air, so you'd be able to pour it into the case if you engineer it correctly. Then seal/cap it, and it would be impossible for water to condense.

I hadn't thought of that- it sounds intriguing, except for two obstacles. First, wouldn't it cost more? Second, the whole reason I want to use CO2 instead of LN2 on the motherboard is to make sure the caps are well ABOVE -150C, the point at which they start to malfunction. With nitrogen, I can't keep track of the temps even if it is gaseous.

Then again I haven't worked with LN2 myself, so I just took some time to try and learn a few things about it...

I appreciate your feedback, I can't even keep track of how many hours I've spent in the past day reading article after article about LN2. I know how you feel...

Also be sure to operate with windows open. If you get a lot of LN2 you could displace all of the oxygen in your room and suffocate. Cheers

Either outdoors- like my driveway- or asking my chem professor in college for sponsorship. It means I'll have unlimited access to both chemicals, as well as an industrial ventilation system. That's not even mentioning professional supervision, which is priceless for a college student who has no experience with subzero temps other than the occasional Jersey blizzard.

Thanks for all your suggestions, keep 'em coming! Any others?

 

[gclg2000]

Could you elaborate on the hard drive not working in a vaccum

 

[mavalpha]

Common problem with sleeve bearings (AKA Fluid Dynamic Bearings, or FDB).

The simplified version: a sleeve bearing is actually an oil-filled tube separating the outer ring from the inner ring, effectively the storage platter from the axle. In a vacuum, the pressure outside is negligible; consequently the tube expands/swells, causing the outside ring, the platter, to stick and not spin anymore.

I don't believe this happens with ball bearings, but most hard drives don't use BB's anymore. I prefer them myself, since I prefer durability at the expense of an insignificant amount of noise. They last longer, and they are less sensitive to environmental fluctuations.

Note that the details aren't 100% accurate, but I wanted to keep it simple.

 

[BrainEater]

Actually , I thought the "not working in a vacuum" issue is related to the read/write heads , and the fact that during operation , the heads are separated from the platter by a thin cushion of air.

Remove the air = Heads touching platter = Dead HDD

 

[zer0signal667]

How about sealing the case with some dessicant inside? No expensive degassing/refilling, no need for vacuum, and relatively low moisture content left in the air.

 

[mavalpha]

Interesting idea, but I'm uncomfortable surrounding my nice, shiny, expensive parts with silica pellets when I don't know how they'll interact with each other.

Vacuuming/purging isn't expensive, the pump itself is $45 on Amazon and the air is free- plus my mom is letting me use hers now. All I need to buy is the dry ice, and the CO2 also helps me cool the motherboard and RAM.

 

[zer0signal667]

Interesting idea, but I'm uncomfortable surrounding my nice, shiny, expensive parts with silica pellets when I don't know how they'll interact with each other.

Vacuuming/purging isn't expensive, the pump itself is $45 on Amazon and the air is free- plus my mom is letting me use hers now. All I need to buy is the dry ice, and the CO2 also helps me cool the motherboard and RAM.

Silica won't harm anything in there, and the pellets usually come in a paper-ish bag for containment anyway. Do you plan on running the vacuum continuously while you have nitrogen and CO2 vaporizing and subliming in there? There's going to be alot of gas trying to escape the case, I would probably not even use a vacuum. All the N2 and CO2 flowing out of there will probably dilute the moisture content down to basically nothing after not too long. And by the way, pulling vacuum over solid CO2 will not keep it solid idefinitely, it would actually increase the sublimation rate.

Also, don't do this in a small enclosed room, remember you need some oxygen left in there to keep living

 

[uclajd]

Nice little experiment. I will subscribe to the thread.

Regarding the Dothan, avoid the P4P800-DX, as they seem to have FSB issues with Dothans. Go P4P800-SE or P4C800-DX.

Check out this thread for more on the Dothan-on-Asus platform.

Good luck!

 

[mavalpha]

Thank you for the heads-up, it gives me all the more reason to stick with my K8's for the project instead of the Banias. If I can dig up a P4C Deluxe for little out-of-pocket expense, I might change my mind.

Also appreciate the interest, and the other suggestions. I'll look into them some more, since I actually have to build the case first- that'll probably take me a few weeks on its own, since summer semester has classes almost every day.

 

[Megadeth_Guy01]

Take your time on constructing the case, if it leaks eek lol. Exploding computer case. err or would it implode?

 

[mavalpha]

Depends. If it's a result of the vacuum, it implodes. If it's from the CO2 gas, it explodes. I'm definitely putting a purge valve on the side to make sure it doesn't explode, but I need to find a cheap guage as well.